Algorithmic Method to Detect Discrepancies in Electronic Medication Histories

Abstract

Medication errors are a leading cause of morbidity and mortality, and impose a tremendous economic and medico-legal burden on society. Discrepancies in medication information across healthcare settings and dispensing pharmacies result in medication errors that are potentially preventable. Method(s) and process(es) using machine-based Boolean logic are described herewith to detect discrepancies in electronic medication histories as effectively as highly trained clinical pharmacists, but with far greater efficiency and parsimony. The reduced cognitive burden and time spent in reconciling discrepant medication information is expected to yield cost savings. Furthermore, if deployed uniformly, the process of reconciling medications electronically may be standardized across Electronic Health Record (EHR) platforms and healthcare settings, resulting in an improvement of medication safety and reduction in medication errors.

What is claimed is: 
     
         1 . A method, when used in a data processing system comprising of at least one processor and at least one memory, bearing instructions to execute operations comprising:
 a. Reading of aggregate medication data, by the data processing system, obtained from computing devices associated with a plurality of different sources and data formats of electronic medication histories for a patient;   b. Analysis of aggregate medication data, by the data processing system, to identify discrepancies in medications by analyzing the content of every instance of the aggregate medication data, further comprising of—
 Determination that every instance of the aggregate medication data, by the data processing system, is indeed a medication as defined by the Food and Drug Administration (FDA) by its standard numeric drug identifiers (NDC/RxNorm—also please see glossary), whereas instances lacking standard numeric drug identifiers are categorized as non-standard discrepancies; 
 Determination that every instance of the aggregate medication data, by the data processing system, is indeed a medication as defined by the Food and Drug Administration (FDA) by its standard numeric drug identifiers (NDC/RxNorm—please also see glossary), whereas instances known to be either an immunization or supply (e.g. glucometer) are identified and categorized as immunization and supplies; 
 Determination that every instance of the aggregate medication data, by the data processing system, is non-duplicative, whereas duplicate instances including partial string matches or evolving signature information (please also see [0003], [0013], and glossary) are identified and categorized as duplicate discrepancies; 
 Determination that every instance of the aggregate medication data, by the data processing system, is non-duplicative, whereas unique (non-duplicate) instances are identified and further categorized as either acute or chronic depending on whether the prescription was issued for greater than 30 days and/or refills authorized; 
   c. Generation of sorted and categorized medication list with discrepancies identified, by the data processing system, obtained from reading and analysis of aggregate medication data from computing devices associated with a plurality of different sources and data formats of electronic medication histories for a patient as outlined above, such that it can be presented within the electronic medication history response in a format that improves medication safety and reduction of medication errors;   
     
     
         2 . A process to improve medication safety and medication management, based on the method of  claim 1 , wherein the operation executed, by the data processing system, analyzes the content of every instance of the aggregate medication data in electronic medication histories, further comprising of—
 a. Generation of sorted electronic medication history response with discrepancies identified in standardized categories of discrepant and non-discrepant medications consistently across EHR platforms so as to promote medication reconciliation and reduce medication errors across healthcare settings. 
 b. Reduce the cognitive burden and time needed to detect and reconcile discrepant medications so as to engage clinicians, and derive efficiencies and cost savings. 
 Furthermore, insights from the method of  claim 1  and the process of  claim 2  may facilitate homogenous exchange of medication information across EHR platforms and healthcare settings through three key domains of healthcare policy—(a) mandates or incentives for importing an electronic medication history prior to medication reconciliation, prescription of new medications, and renewal of chronic prescriptions, (b) mandates or incentives for information exchange across all electronic prescription networks including ‘self-contained (closed)’ networks such as Kaiser Permanente (Gabriel and Swain, 2014), and (c) mandates or incentives for the universal adoption of standard contemporary numeric drug identifiers, namely RxNorm (see glossary) that is continuously updated and maintained.